Zachary Christensen and Dr. Michael Findley, Department of Political Science
Ugandan non-governmental organizations (NGOs) have played a prominent role in the transmission gender equality norms in Ugandan. Uganda typically fares poorly in international measures of gender equality which in part seems to stem from the socially conservative political views of the country. Notwithstanding those views, Uganda has recently made significant progress towards implementing more progressive legislation regarding women. The Domestic Violence Act passed by the Parliament of Uganda with backing from the President was one piece of evidence of this shift, establishing firm punishment for domestic violence. The government has also taken increasingly aggressive measures to end female genital mutilation. Ugandan civil society organization and NGOs have been active lobbyists for these changes.
The international aid community played an active role in funding and encouraging these organizations to promote gender issues. If donors were facilitating more action about widely-held views, these actions are laudable. However, if NGOs and their workers do not truly believe in gender equality (at least as presented by Western aid agencies), such funding may ultimately be wasteful or could unduly manipulate the domestic politics of the country by world powers.
In this study, I sought to assess attitudes on gender among Ugandan NGOs. I distributed a survey via email to approximately three hundred NGOs in Uganda in late March 2011. 146 organizations completed the survey, many of whom were from organizations that have frequent interaction with international donors.
77% of the respondents reported that their organization experiences difficulty in getting funding. 80% reported that obtaining funding was the biggest challenge their organization faced. In such an environment, respondents may feel pressure to mask their true feelings about gender projects and pursue the grant funding available for such projects. 47% of the survey’s respondents reported that they had done gender-themed projects. Two clear messages emerge, gender is a common project area and NGOs have a strong need for good relationships with donors.
Donors may not be pleased to hear that NGOs do not agree with the donor agenda; therefore, if asked directly, NGO workers would have reasons to mask their attitudes on the subject. To reduce such a problem, a list experiment or item count survey question was used. Instead of asking NGO workers if their organization had incorporated gender equality internally, we asked them to look at a list of statements about their organization and report how many items were true but not to disclose which items. This way, the surveyor has no way of knowing which specific item was selected, only how many. In such a situation, the respondent could plausibly deny having admitted to something as sensitive as being gender biased. To determine attitudes about the sensitive item, survey respondents are randomly assigned to receive either a control list or a treatment list. Both lists would contain several items, but only the treatment list would contain the sensitive item. Using a simple difference in means test, attitudes about the sensitive item can be evaluated.
The question was as follows:
Please tell us how many of the following statements are true about this organization? Please do not say which statements are true, only how many.
- Interactions with local government officials have helped this organization operate more efficiently.
- This organization is well-funded.
- There is real gender equality within this organization.
- This organization contributes to development.
- This organization would benefit from an increase in volunteers.
- This organization seeks to promote the interests of the national government.
The sensitive item in the list, “There is real gender equality within this organization”, asks a central question about the moral authority with which Ugandan NGOs can advocate gender programs. If many of the respondents feel that their organizations do not have gender equality the CSOs risk being perceived as duplicitous. This question is sensitive because it calls for introspection and potential self-criticism. For social reasons and financial reasons, organizations may not want potential funders and the broader community to question the internal problems they face. The list is designed to minimize this effect by reducing allowing the respondent the option of deniability.
Results
Using the package “list” developed by Kosuke and Imai (2010) for R software, the first analysis provides a simple difference-in-means analysis of the lists. On average, the treatment list had 0.66 more items than the control list, with a high level of statistical significance (p <0.01). In other words, many of the respondents admitted that their organization did not have gender equality. To explore what can help determine whether or not an organization has gender equality problems, we employed a multivariate regression analysis based on other information gleaned in the survey. We tested for the respondent’s gender, whether or not the organization has participated in gender projects, whether or not the organization is based in rural areas, and the level of funding they have received from foreign donors. Surprisingly, none of these factors is sufficient to provide a statistically significant explanation for gender inequality. Two questions were meant to detect the ability of donors to screen for gender bias. One question measured the difficulty of the organization to find funding (as reported by the NGO). The second question asked how many grants the NGO had received in excess of US $10,000 in the last year. These variables failed to explain the likelihood of the survey respondent of reporting that there was gender inequality in the organization. This was true even when there was a control for whether or not the organization was actively executing gender projects. This is surprising. NGOs in Uganda have made some political progress in promoting gender equality. Certainly, this could have positive effects. However, this survey indicates that gender equality remains a real struggle within Ugandan NGOs themselves. This survey was unable to develop a clear explanation about what demographic or organizational characteristics seem to drive the gender problems within organizations. A follow up question was added to the survey to more thoroughly explore the relationship between donors, NGOs and gender attitudes. However, robustness tests for the survey results on that question detected a survey sampling bias. Additional research should be conducted to better understand what prevents these organizations from attaining gender equality.